Corso di Dottorato di ricerca
in Scienze dell’Antichità
CICLO XXXI
Tesi di Ricerca
Textile tools from Egypt
and Southern Levant
SSD: L-OR/05; L-OR/02
Coordinatore del Dottorato
ch. prof. Luigi Sperti
Supervisore
ch. prof. Lucio Milano
Dottorando
Chiara Serena Spinazzi-Lucchesi Matricola 987987
1
Summary
1Introduction ... 5
2Raw materials ... 11
2.1 Flax... 11
2.1.1 First flax textiles... 12
2.1.2 Fibre preparation ... 16
2.2 Hemp ... 17
2.3 Nettle and ramie ... 17
2.4 Cotton ... 18
2.5 Wool ... 19
2.5.1 First woollen textiles ... 19
2.5.2 Wool processing ... 22
3From fibre to textile: spinning and weaving ... 24
3.1 Spinning ... 24 3.2 Spinning tools ... 30 3.2.1 Spindles ... 30 3.2.2 Spindle whorls... 31 3.2.3 Perforated sherds ... 34 3.3 Weaving ... 36 3.4 Weaving tools ... 40 3.4.1 Loom weights ... 40
3.4.2 Spatulae and pin beaters ... 43
3.4.3 Sewing needles ... 44
3.4.4 Combs ... 45
4Textile tools from Southern Levant ... 47
4.1 Spindles ... 47
4.1.1 Spindles from Southern Levant... 51
4.2 Spindle whorls... 53
4.3 Perforated sherds ... 64
4.4 Spinning bowls ... 65
4.5 Bone spatulae and beaters ... 68
4.6 Basalt rings ... 70 4.7 Loom weights ... 74 4.8 Needles ... 80 4.9 Megiddo ... 83 4.9.1 Introduction ... 83 4.9.2 Spindles ... 86 4.9.3 Spindle whorls... 87 4.9.4 Perforated sherds ... 91 4.9.5 Spinning bowls ... 92 4.9.6 Loom weights ... 92 4.9.7 Basalt rings ... 95
2
4.9.8 Bone spatulas and beaters ... 96
4.9.9 Needles ... 97
4.9.10 Contexts of findings ... 97
4.9.11 Conclusions: development of textile tools at Megiddo ... 98
4.10 Beth Shean ... 101 4.10.1 Spindles ... 104 4.10.2 Spindle whorls... 104 4.10.3 Perforated sherds ... 108 4.10.4 Spinning bowls ... 109 4.10.5 Loom weights ... 110 4.10.6 Basalt rings ... 112
4.10.7 Bone spatulae and beaters ... 113
4.10.8 Needles ... 113
4.10.9 Contexts of findings ... 114
4.10.10 Conclusions: development of textile tools at Beth Shean 115 4.11 Hazor ... 118 4.11.1 Introduction ... 118 4.11.2 Spindles ... 121 4.11.3 Spindle whorls... 122 4.11.4 Perforated sherds ... 124 4.11.5 Spinning bowls ... 126 4.11.6 Loom weights ... 126 4.11.7 Basalt rings ... 129
4.11.8 Bone spatulae and beaters ... 129
4.11.9 Needles ... 131
4.11.10 Conclusions: development of textile tools at Hazor 132 4.12 Tell el-Far’ah (N) ... 134
4.12.1 Introduction ... 134 4.12.2 Spindles ... 137 4.12.3 Spindle whorls... 138 4.12.4 Perforated sherds ... 141 4.12.5 Spinning bowls ... 141 4.12.6 Loom weights ... 141 4.12.7 Basalt rings ... 143
4.12.8 Bone spatulae and beaters ... 144
4.12.9 Needles ... 146
4.12.10 Conclusions: development of textile aools at Farʽah 146 4.13 Conclusions: Levantine textile tools ... 149
5Textile tools from Egypt ... 152
5.1 Spindles ... 152
3
5.3 Perforated sherds ... 166
5.4 Spinning bowls ... 168
5.5 Bone spatulae and beaters ... 171
5.6 Basalt rings ... 174
5.7 Loom weights and net weights... 174
5.8 Needles ... 178 5.9 Kahun ... 182 5.9.1 Introduction ... 182 5.9.2 Spindles ... 185 5.9.3 Spindle whorls... 187 5.9.4 Spinning bowls ... 191 5.9.5 Weights ... 192 5.9.6 Needles ... 193 5.9.7 Spatulae ... 193 5.9.8 Heddle jacks ... 194 5.9.9 Fibres ... 195
5.9.10 Textual references on textile activities ... 196
5.9.11 Conclusions ... 197
5.10 Gurob ... 199
5.10.1 Introduction ... 199
5.10.2 Foreigners at Gurob and textual sources ... 201
5.10.3 Spindles ... 204 5.10.4 Spindle whorls... 206 5.10.5 Spinning bowls ... 211 5.10.6 Weights ... 211 5.10.7 Needles ... 212 5.10.8 Spatulae ... 212 5.10.9 Fibres ... 214
5.10.10 Other tools connected with textile production 215 5.10.11 Conclusions ... 216
5.11 Deir el-Medina ... 218
5.11.1 Introduction ... 218
5.11.2 Tools from Bruyère’s publications ... 219
5.11.3 Tools of the Museo Egizio collection ... 222
5.11.3.1 Spindles ... 223
5.11.3.2 Spindle whorls... 226
5.11.3.3 Wood identification... 232
5.11.3.4 Needles ... 234
5.11.3.5 Fibres ... 235
5.11.4 Textual references on textile activities ... 236
5.11.5 Conclusions ... 237
5.12 Conclusions about Egyptian textile tools ... 240
4 7Bibliography ... 250 8Catalogue ... 279 8.1 Megiddo ... 280 8.2 Beth Shean ... 362 8.3 Hazor ... 406 8.4 Tell El-Far’ah (N) ... 449 8.5 Kahun ... 484 8.6 Gurob ... 508 8.7 Deir El-Medina ... 536 8.8 Other Sites ... 571
5
1 Introduction
The aim of the present work is to compare the developments in the field of textile production technology in the southern Levant and Ancient Egypt. Notably, its main concern is to understand to what extent the two regions had a shared tradition or, conversely, if they take different directions from a certain moment onward. As a matter of fact, many contributions mostly focus on comparing the two textile traditions, without directly investigating the tools and techniques used in both areas. The issue in such studies lays in the scarcity of published tools, which has been the cause of limited knowledge in the actual industry of spinning and weaving. Such a lack of data has, as a major consequence, lead to incomplete or incorrect conclusions in the field. Therefore, the present work proposes to amend such lack in the knowledge of textile production by directly analysing materials from selected sites of both southern Levant and Egypt. Nonetheless, also by using all the documentation available from published excavations reports the present contribution intends to provide a picture of the developments and an easier and more clear way to compare these two regions. The broad geographical area covered by the research comprehends Egypt, from Delta to the Nubian border (with also a few sites slightly southern considered for the importance of their findings) and sites in the Southern Levant, such as Israel, Palestine and the Jordan western area.
The chronology is extremely wide as well, since it takes into considerations tools from the Neolithic to the beginning of the Persian period. After that period, reports of a decrease in loom weights weight in the Levant might points toward the introduction of new fibres, as suggested also by textual sources. Therefore, it appears compulsory to investigate the second half of the first millennium with a more focused study that, at the present state of the archaeological documentation, was not possible in this work. During the whole period under investigation, few fibres were used for textiles production, mostly flax and wool, and changes in tools could be linked to specific textiles productions rather than with the adoption of a new fibre.
The objects investigated in the present work do not permit the creation of a specific and intrinsic chronology. The study has, therefore, to rely on information provided by excavators and on the updating of sites stratification. Also, given the broad chronology and geographic area under study, the thesis was not focused on micro-chronological patterns of change but aimed to understand picture of textile production as a whole. For this reason, discussions concerning chronological reassessments between the two regions, especially for the EB II – Early Dynastic and for the passage from the Late Bronze period to the Iron Age, are not covered here. However, the objective problems currently under debate are considered and taken into account (Mączyńska 2008; van den Brink and Levy 2002; Stevenson 2016; Finkelstein et al. 2017). Different terminologies used by various authors have
6 complicated the picture, too. Consequently, a simplified chronology with some of the main alternative names is adopted as presented in tables 1-2.
Table 1 Chronological chart of Ancient Egypt1
Upper Egypt Lower Egypt
Fayyum A (5400-4300) El-Omari (4700-4300)
Badarian 4400-3800 Tasian? Merimde (4700-4200)
Naqada IA-IIB 3800-3450 Predynastic Buto I/Maadi (4000/3900-3500/3400) Naqada IIC-D 3450-3325 Naqada IIIA-IIIB 3325-3085
Naqada IIIC-D/First dynasty 3085-2867
Early Dynastic period/Protodynastic (Dyn. 1-3) 3000-2500
Old Kingdom (Dyn. 4-6) 2500-2110
First Intermediate Period (Dyn. 7-10) 2110-1980
Middle Kingdom (Dyn. 11-12) 1980-1760
Second Intermediate Period (Dyn. 13-17) 1760-1540
New Kingdom (Dyn. 18-20) 1540-1077
Third Intermediate Period (Dyn. 21-24) 1076-723
Late Period (Dyn. 25-30) 723-343
Table 2 Chronological chart of the Levant (Sharon 2014)
Period Alternative names Absolute date
9750
PPNA 8500
PPNB 7750
[PPNC] 6500
Ln1 PNA, Early pottery
Neolithic 6000
Ln2 PNB, Late PN, Early Chalcolithic 5500
[Middle Chalcolithic]? 4500
Chalcolithic Late Chalcolithic 3900/3700
EBI Proto-Urban 3200/3000 EB II 2850/2600 EB III 2500/2300 IB MB I (Albright) EB IV (Syria) 2200/1900 MB I MB IIA MB I 1750 MB II (early) MB IIB MB II
MB II (late) MB IIC MB III 1640/1540
LB I LB I LB I 1400
LB IIA LB II 1330/1300
7
LB IIB LB III (Amiran) LB II 1200/1150
LB/IR IA IA LB III
(Panitz-Cohen/Ussishkin) 1150/1000
Iron I IA IB
[Iron I/II] (Mazar) Early IA IIA 1000/900
Iron IIA IA II 925/800
Iron IIB IA III IA IIB-C 730/700
Iron IIC IA III IA IIB-C 608-586
Iron III/Persian IA III
All the objects that could be possibly related to textile production are included in the study, from traditional spindle whorls and loom weights to objects with a less certain function, such as perforated sherds, basalt rings and pin beaters. The purpose is to understand when they appear in both areas and if they are particularly spread in specific periods and less in others. The inclusion of more controversial and less traditional objects in the research allows a deepened analysis of their manufacture, the presence or absence of wear traces and their finding in connection with other objects linked to textiles.
In the choice of the categories of objects, the most recent studies about textile tools have been followed. Firstly, studies directly connected with the areas under exam have been helpful to understand which were the problems under debate and which sites could provide the most useful sets of tools to resolve them. The starting point is “Prehistoric textiles” of E. Barber, which present an overview of some of the most important findings in the Mediterranean area and still represents a fundamental instrument of research. Concerning the Near East, a few more specific studies have appeared, especially “Gli strumenti di tessitura dall’Età del Bronzo all’Epoca Persiana” of L. Peyronel, “Essai sur le tissage en Mesopotamie” by C. Breniquet and “Textiltechnik im Alten Orient” by E. Völling. In addition to the mentioned indispensable works, the publications and studies of O. Shamir must be taken into consideration. Although she has not yet presented an overall work on textiles production in Israel, her studies on textiles and tools carried at dozens Levantine sites are nonetheless, a needful instrument for the study of the region Excluding the work of B. Kemp and G. Vogelsang-Eastwood's “The Amarna textile industry”, only a minority of articles and books deal with textile production and even less with tools.
From a methodological point of view, the publications of the Centre for Textile research in Copenhagen have provided the parameters for studying tools, like in the case of minimum or maximum weight of a spindle whorl or a loom weight. Their work is mostly based on their personal experience and research, coupled with a strong programme of testing techniques and ancient tools to grasp them as better as possible. Given the different sources of fibres in Northern Europe, mainly
8 wool, and South Mediterranean, where flax is predominant, not all the results and parameters can be directly applied to the areas under study and will be therefore discussed in light of the actual findings from Egypt and Southern Levant.
For acquiring and refining the necessary skills for studying fibres and textile tools, and for
understanding their usage, was necessary to receive a specific training; in part previously gained before starting the PhD. Throughout the three years, an intensive course of weaving was performed in Turin, by Marina Costantino, a local weaver. It was an individual face to face lesson structured around the specific needs of the author, focused on acquiring basic knowledge about different weaving techniques and understand all the processes of setting a loom. Furthermore, card weaving was also included.
9 A second, more complete course was attended at the Textile Research Centre in Leiden. It was structured in several sections aimed to give each a broad knowledge of textiles. The course analysed different types of fibres, wool, flax, silk, cotton as well as less common fibres, tested their characteristics and analysed them using an optical microscope. The course included sessions of spinning and weaving and parts of the fibres were dyed. Finally, it analysed different types of textiles, from the Neolithic remains of charred linen fabrics from Çatal Hoyuk to modern silk kimonos, to teach how to recognise fibres and weaving techniques by observation. The course was taught by G. Vogelsang-Eastwood, the highest expert of Ancient Egyptian textiles.
Finally, during a stay at the Centre for Textile Research in Copenhagen, it was possible to spend a day at the Open Air Museum of Lejre where ancient textiles are recreated with the same tools and substances used in antiquity. It was possible to carry on a simple
Figure 2 Attempts of spinning with a spinning bowl in Leiden
10 experiment which consisted in preparing flax fibres for spinning. After retting (which requires several days and was not performed that day), flax fibres were subjected to breaking, scutching and hackling (combing with iron teeth combs which were not available for periods under study), spliced and spun. Given the broad chronological and geographical setting, this thesis will not provide a list of all the textile tools published from Egypt and Southern Levant. Instead, it focuses on case studies and compares them to relevant materials known from the two areas. Originally, nine case studies were selected, but problems concerning permissions for studying of materials and difficulty of locating them have restricted the work to seven sites. Sites selected are Megiddo, Hazor, Beth Shean and Tell el-Far’ah (N) for Southern Levant, while for Jericho and Tell el-Far’ah (S) only few items were available and will not constitute a case study. Kahun, Gurob and Deir e-Medina are the three sites selected for Egypt; however, it seemed relevant for this study to locate some Pre- and Proto-dynastic materials, generally poorly published, to better understand the development of the Egyptian textile industry. Materials from these sites will be listed in the catalogue under the entry “other sites”. Each case site is studied individually and the most relevant findings are further discussed in the region general chapter.
The thesis is therefore structured as follows: a first introductory chapter about fibres and first textiles findings regarding the two areas under study, so to provide a frame of possible raw materials and techniques available for the periods under examination; a second chapter about spinning and weaving techniques and the presentation of categories of selected tools. The core of the thesis is made up of two chapters, one for Southern Levant and one for Egypt. Each chapter discusses the different categories of materials (e.g. spindles, spindles whorls, etc.) followed by subchaptersrelated to case studies. At the end of each chapter, conclusions about the area examined are provided. The last chapter is dedicated to the final conclusions and to comparison between the two regions. It is then followed by a catalogue of all materials studied and by drawings of selected items.
11
2 Raw materials
In this chapter, a wider geographical area will be considered to understand which types of fibres were actually available in Egypt and Southern Levant during the long period under study. Evidence from the Mediterranean area, Mesopotamia, Iran and India will be therefore included, but the focus will be maintained on the geographical area under examination.
2.1 Flax
There are around 230 species of Linum around the world, but the most common wild flax is Linum angustifolium: it is widespread in Mediterranean and most Europe and produces blue flowers, while other sub-species are characterized by pink, yellow and white flowers (Baines 1989, 13). Linum angustifolium is probably the wild ancestor of the domestic Linum usitatissimum, from which the fibre used for textile production is extracted. According to R. Forbes (Forbes 1956, 4:27), the wild variety grows in the Mediterranean and Atlantic coastal areas as well as in Iran and Iraq, but not in Egypt.
Flax is the term used to denote the plant and its products, seeds and fibres, while linen is the final product of the process, which converts fibres into a spun thread. The stem of flax is composed by a central void, called lumen, and a woody core, around which the fibres, known as bast, protect and support the central part where nutrients are carried, and closed by a cortex and
an epidermis. Fibres occur in bundles, extended from the root of the plant to the tip and are glued into a substance composed by pectineus gums, waxes and non-cellulosic materials, which should be removed in order to prepare the fibres for spinning. Each stem can contain a variable number of bundles, generally between 15 and 35, and each bundle can contain from 10 to 40 single fibres or ultimate bundles; the ultimate bundles generally remain together during the transformation into linen (Kemp and Vogelsang-Eastwood 2001). They are three times thicker at the root end than at the top. The main component of fibres is cellulose, with small cells, which overlap at regular intervals. A
Figure 4 Linum
usitatissimun (Kohler
Medizinal Pflanzen)
Figure 5 Section of a stem of the flax plant (Kemp/Vogelsang-Eastwood 2001, fig. 2.1)
12 single fibre can measure from 45 to 100 cm and its appearance is smooth, cylindrical, with joints along the length, where cells overlap.
Flax fibres have a natural twist to S, which becomes visible when they are wet and held at one end; the other end will twist to S naturally, due to the orientation of the microfibrils inside the fibres.
2.1.1 First flax textiles
Flax produces both fibres for textiles and seeds useful for eating and producing oil and its stems may have been used in the first times to produce also basketry or cordage, because of its great strength. It is not known which of these products encouraged men to domesticate it, but it certainly happened in very early times. Recent findings of wild flax fibres in Georgia date its first usage at least to 30000 years ago (Kvavadze et al. 2009).
The first thread in the world is made of flax and comes from Wadi Murabbaʽat in the Judean Desert. It was not found in a regular excavation but it was purchased from Beduin, who plundered the caves; nevertheless, the analysis made with an Accelerator Mass Spectrometry prove that it dated to 10.220±45 BP. It is wound around wooden pegs in order to fasten them and
to create a sort of wooden comb. It is also covered by asphalt. Yarn is z-spun and S-plied and build a twine structure which probably originally covered most of the comb (Schick 1995, 201).
The earliest examples of linen fabrics come from the well renowned Na al emar cave in the Southern Judean Desert and are dated to the Pre-pottery Neolithic B (6900-6300 B.C.). In this cave hundreds of artefacts made of vegetable fibres (including Linum sp.) were preserved, like cordage, baskets, mats, yarn and fabrics (produced with different techniques like looping, knotting, twining and weaving) (Shamir 2014 p. 143). Cordage is generally produced with unspun strands, with a final twist to Z, while nets are made of linen threads z-spun and S-plied, as are the majority of yarns found in the cave. Some specimens are s-spun and Z-plied and a few items are cabled. Fabrics are not actually textiles, since they are not woven but others technique like knotted netting and knotted looping are employed, but the most frequent technique is weft twining. Actual textiles fragments are probably later in time, since C14 dated them to 660±200
Figure 6 Weft twining example (Schick 1988:37)
13 years BP (Schick 1988, 38). Several needle-shuttles and awls were also presents, but no other tools directly connected with textiles production (i.e. whorls or parts of looms).
Actual textiles were found in the Pre-pottery Neolithic levels of Çatal Höyük. They have been preserved thanks to high temperatures caused by the great fire, which destroyed buildings of Level VI. Under houses and shrines, clay layers sealed several tombs, which preserved several fragments of textiles wrapped around bodies. There are both twined fabrics and tabby weave textiles, with different density in weave and different thickness of threads. Threads are z-spun and S-plied (Burnham 1965, 171), as the majority of Na al emar. There has been a large debate over the identification of fibres in past years since at the beginning they were said to be wool. Their brittle and charred conditions result in difficulty in seeing fibres characteristic under a polar microscope. Helbaek (Helbaek 1963, 44) was the first to conduct studies on this fibres, without using sophisticated analysis, and concluded that most of them were made of wool, while some strips were made of vegetable fibres. Microscope analysis conducted subsequently by Burnham (Burnham 1965, 170) identified the presence of scales on the surface but the absence of kemp and hair in textiles made the author hesitate about this identification. Presence of wool was completely ruled out by Ryder (1965, 175; Ryder and Gabra-Sanders 1987, 91), which conducted several tests on the fibres as well as SEM analysis. Flax was the most probably fibre used for producing Çatal Höyük textiles and it was confirmed by recent findings. Several new fragments of textiles were found in new excavations in the site and those submitted to SEM analysis were proved to be Linum usitatissimum (Fuller et al. 2014, 121–22). As the older findings, threads are z-spun and S-plied (S, 2z).
Some evidence of flax in PPN levels have been found also in the Anatolian site of Çayönü, but organic materials are preserved only carbonized or mineralized. A pseudomorph fragment of cloth has been found wrapped around a tool made from an antler and has been dated to 7000 BC. Fibres were not spun but twisted together as in twine and set on a frame made of four sticks, but not produced on an actual loom. Also, linseeds were found in Çayönü and they were recognised as wild flax, probably Linum bienne, but certainly not Linum usitatissimum, which instead is attested in Tell Ramad (Syria) around 7190-6700 BC (Zeist and Roller 1991, 82).
Jarmo, in northern Iraq, provided information about weaving skills but unfortunately not on raw materials. In this site, in fact, impressions in clay and bitumen of textiles and basketry have been preserved and among textiles, tabby and basket weave are present. All these impressions are dated to 7000-6000 BC (Adovasio 1975, 224).
The Late Chalcolithic period (4500-3700 BC) has preserved a larger number of linen textiles, especially from caves in the Judean desert in Israel. These caves were used as refuges, burial places and sometimes for living in the Chalcolithic period, for a short time in EBI (Davidovich, 2012) and
14 again in the Roman Period. Textiles from the earlier period are quite different from those of the Roman period, both for fibres and techniques. Wool, for instance, seems not yet be used and thread production is characterised by splicing. Several caves have yielded linen textiles, but the most famous are the cave of the Treasure, Nahal Mishmar (a burial site rich of textiles, basketry and cordage) and the Cave of the Warrior2. From this evidence it is clear that Chalcolithic threads are generally s-spun and S-plied (s, 2S) with clear evidence of the use of splicing; weave is generally a simple tabby, sometimes warp-faced with very few decorations made by hollows and soumak. No dyes have been so far recognised, but black stripes made of paint or asphalt are visible. Flax was probably cultivated in the Jordan Valley, at sites like ʽEn Gedi, Jericho or the Beth Sheʽan Valley and not imported from elsewhere, like Egypt, (Shamir 2015, 21), but no specific studies of fibres provenance have been conducted. It is inferred from the very different quality of the contemporary Egyptian textiles, but since the paucity of evidence of the Egyptian production in this period, it seems difficult to make such a strong statement.
Egypt, in fact, has preserved a very limited number of archaeological sites for the period between 7000 and 5400 BC (Shaw 2003, 35) and, so far, no traces of textiles have been found. The first Egyptian textile comes from the Neolithic Fayum A culture (around 5000 BC3) together with evidence of basketry. It is a tabby textile made of linen with loosely spun yarn with a z-twist and S-ply4 and an uneven woven (Jones, 2008, p. 107). It was found in upper K pits, a large series of Neolithic granaries of the Fayum, northeast of Kom W. The linen piece was inside a disintegrate cooking pot and was found loose in the gravel outside the annex granary to silo 16 (Caton-Thompson and Gardner 1934, 46) together with two oval polishing pebbles and a fish vertebra. Analysis initially did not prove it was made of Linum usitatissimum; however, in silo 59, a flax seed of Linum usitatissimum was
2 For a complete list of caves and textiles see Shamir (2015) and Schick (2002).
3 New radiocarbon tests of the Upper K Pits have confirmed such a date, although with a very large errors (Wendrich et
al. 2017, 120–21). No other textiles have been found, but several specimens of basketry are attested.
4Schick (1988, 40–41) reports a personal communication of R. Hall (curator of the Petrie Museum) which states that the
Fayum textile is s-spun and Z-plied, differently of what stated by Crowfoot (1958, 431), who wrote z-spun and S-plied. Jones definitely proved it is z-spun S-plied.
15 recognised. A recent analysis conducted on this textile have definitely confirmed the original identification (Jones and Oldfield 2007, 34).
Several textiles have been found in the Badarian tombs but they occur less frequently than skin garments (Brunton & Caton-Thompson, 1928, p. 40). It is an interesting detail since the following tradition will tend to exclude the use of animal skins and fibres from the funerary ritual. These skin garments were more frequently worn by men rather than women and hair is inside, in contact with bodies. Generally, they are goat and antelope skins, but, at least in two cases, a black fur has been recognised. Linen garments, shrouds and bandages are present, but less frequently and in bad condition. They can occur in the same burials of skins and they can also cover the head of the deceased, while skins seem not to occur higher than shoulders. Unfortunately, these textiles are in such a bad condition that it is difficult to recognize the fibre of which
they are made. In his reports, T. Midgley states that they are probably not made of linen but of a vegetable fibre like grass (Brunton & Caton-Thompson, 1928, p. 67). Flax was known in these sites since one flax seed has been recognised5. Recent analysis conducted by J. Jones (2002, 326), however, are questioning the identification of fibres in early analysis and it is probable that most of the Egyptian fibres were actually flax. One peculiarity of Badarian textiles is that warp and weft are not at a right angle, but weft yarns are put diagonally (with an angle which can vary between 20-40°) to avoid that the textile frays. The same characteristic is present in some Predynastic textiles but seem to disappear from Naqada III onwards (Jones, 2008, p. 113). It has also been suggested that a change in the weaving equipment occur in the Predynastic period, but it has not been recognised yet.
A greater number of linen textiles survived in Egypt from the Predynastic period6 onwards allowing a better knowledge of Egyptian textile production than that of every other ancient culture. However, it has always to be kept in mind that Egyptian textiles come nearly exclusively from tombs and this fact restricts the typologies of cloth
survived and limits our knowledge of the actual dresses and fabrics used in daily life.
5 Badari tomb 3000, object 3, identified as Linum bienne (Brunton & Caton-Thompson, 1928, p. 63) 6 For a list of textiles recently analysed see (Jones, 2008, p. 101)
16
2.1.2 Fibre preparation
Linen is a very suitable fibre for clothing, but also for curtains, bed sheets and towels. It has several properties, being comfortable, fresh and non-itching; it can be washed several times without worn and tend to become softer after a long use. It is easy to clean and to remove dirt and its transpiring property make it a perfect choice for warm and dusty countries. All these qualities well explain its success in ancient times, in spite of the difficult and time-consuming process to make it. Egyptian representations show just part of this process, the moment of harvesting and the preparation of yarn for spinning, leaving in doubt how exactly fibres were extracted from bundles. Generally, the process includes plucking the plant, beating or rippling it to eliminate the seed heads, retting, drying and beating or scutching and finally hackling the stems7. The ancient system was briefly but exhaustively described by Pliny8. Stems are dried for a few days and free from the seed heads by beating with a wooden mallet or with the aid of a wooden comb. To separate stems from fibres, retting is the method usually practised in ancient times (but no evidence are available for Ancient Egypt)9. They can be either submerged in a flowing stream or in low pools or left in the field and exposed to the action of dew and rain. The aim is to dissolve the pectinous substance, which surrounds the fibres, by fermentation. The first method is certainly the faster and produce a golden fibre, while the second fibres of a darker and greyish colour.
At the end of retting, the fibres have to be dried again and clean from the bark and woody core and it can be done with sticks (willowing), mallets or using a wooden knife which scrapes the fibres (scutching). Finally, fibres have to be combed (hackling) to separate the long fibres, suitable for spinning, from the shorter (tow), which can be used either to produce a coarser yarn or as padding for cushions and mattresses.
Flax can be cultivated both for seeds and fibres, but it is not possible to obtain both seeds and a good quality of fibres. For seeds, in fact, plants should have more space among each other in order to have a shorter plant but with more branches. For textile production, the longer the fibre the better it is. To obtain long stalks, seeds have to be sown near each other so the plant will grow taller with branches on the top. Furthermore, the moment of harvesting will influence the final result of the fibre. Young and green plants will give finer and better quality fibres than ripe and yellow plants. To preserve the entire length of the plant, it should not be cut but plucked.
7 Several authors describe this process, e.g. Forbes (1956, pp. 30), Barber (1991, p. 13) and Peyronel (2004, pp.
29-31), more specifically Baines (1989, pp. 13-14) and Kemp/Vogelsang-Eastwood (2001, pp. 25-34).
8 See Forbes (1956, p. 29) for description and comment.
9 See Vogelsang-Eastwood for a discussion about retting in tombs paintings and the possible use of decortification (Kemp
17
2.2 Hemp
Hemp (Cannabis sativa) is another bast fibre, which has stems longer than flax (high up to 5 m) and from which coarser fibres and textile can be produced. It is very suitable, thanks to its strength, to produce ropes and cords and for sails because of its resistance to seawater (Peyronel, 2004, p. 31). If the association of Fiber A with hemp in Midgley reports (Brunton, 1937, p. 145) is correct, then it is highly probable that it was used in Badarian, Predynastic and pan-graves cloths10.
Other ancient findings from Europe date its first usage to 5500-4500 B.C., but they consist only of seeds and not of actual fabrics (Peyronel, 2004, p. 32). Since it has a very similar structure to flax, it is very difficult to recognise it correctly. They differ because flax has a natural S-twist, while hemp a natural Z-twist; it means that the fibre, once wet and stuck on one end, will start to twist at the other end following its natural inclination. More recent and trustworthy analysis have evidenced that hemp was certainly in use in the Levant in the Chalcolithic phase. It was used a DNA analysis and it proved that not only it was used to produce textiles, but also that a mixture of hemp and flax was employed, making the identification of the two fibres, without sophisticated analysis, extremely difficult (Murphy, 2011)11. The process of preparation of fibres is very similar to that of flax.
2.3 Nettle and ramie
Nettle (Urtica dioica) might be employed in textile manufacture; nettle stems contain bast fibres and they are extracted with a similar process as flax, but are coarser. It might be used to produce cloth as well as sacks and ropes. Its use as textile and basketry fibre might be very ancient, if the imprints found from the Czech Republic at Dolní Vestonice, and which are dated to 32-29000 years ago, are really of Urtica sp. (Soffer, 2001).
Another member of Urtica family is Boehmeria nivea, from which the more used ramie is obtained. Nettle fibres have been positively recognised in a textile from Kastelli Chania in Greece and it is dated to the Minoan period. The fragment of fabric is a dense tabby weave, weft-faced, decorated with an additional thread. Warp is made of S-plied thread, probably of flax fibres, while weft consists of un-plied threads of animal origin, probably goat hair. The additional thread has been recognized as being S-plied, of plant origin, and its characteristics make highly probable an identification with nettle fibres, more specifically ramie (Moulhérat and Spantidaki 2009). The use of goat hair and ramie
10 See Barber (1991, p. 15) for another proposal of identification as Hibiscus cannabinus.
11 Other methods of fibres differentiation are the presence of calcium oxalate cluster crystals in nettle and hemp, but not
in flax or the analysis through synchrotron radiation micro-beam diffraction and micro-fluorescence. See Gleba (2011, p. 9).
18 in the Minoan period is very interesting because it proves the continuity of usage of a larger number of fibres besides the more common flax and sheep’s wool known by textual sources.
2.4 Cotton
A very intriguing fibre is cotton. It is native of India, Senegal and Sudan where first evidence of this plant has been found. It does not belong to the bast family, but it is derived from many trees and herbs of the Gossypium family. Fibres are very short and grow on branches as padding for seeds. It is generally thought that cotton arrives in the Mediterranean area from the Indus Valley, where it was used to produce textiles in Mohenjo Daro and Harappa at least from 2500 B.C. onwards (Peyronel, 2004, p. 33). After these first examples, cotton disappears from Mesopotamia or Levant to reappear only in the first Millennium BC; garments made of cotton fibres have been found in the Assyrian Queen’s Tomb in the 8th century BC. Also, texts seem to indicate the exploitation of cotton during the first millennium. The Akkadian word kitinnû and the Neo-Babylonian term karpassu have been proposed to be identified as cotton, as well as biršu in Assyrian texts. The first mention of cotton trees appears in the royal text of King Sennacherib (704-681 BC) which quotes a “tree bearing wool”. However, the origin of these trees is indicated as Babylonia, so this fibre could have been known in that region before it arrives in Assyria (Gaspa 2017, 157–58).
Herodotus writes that a corselet made of cotton was sent from pharaoh Amasis as a gift to Athena (Barber, 1991, p. 33), but no findings of cotton are known from Egypt in this period.
First archaeological evidence of cotton comes, surprisingly, from Jordan, where the fibre is not native. The site of Dhuweila in eastern Jordan have provided, in fact, some textile impressions on plaster but also small tufts of fibres, which have been radiocarbon dated to 4500-3000 BC, i.e. Levantine Chalcolithic or Early Bronze Age phase. Little doubt can be raised about the correct identification since it has been observed under a Scanning Electron Microscope, which clearly shows the typical convolutions of cotton fibres. The problem is the origin of the fibre, which was certainly imported since it is not native of that region. Both India (Mohenjo Daro) and Sudan (A-group culture of Nubia) cotton samples are dated to the mid/late-3 millennium B.C., but the findings from Mohenjo Daro, associated with materials of the mature Harappan Period, might be later. Only the Nubian materials have a chronology near (but lower anyway) to that of Dhuweila. Since no findings of cotton (seeds or cloth) are known from Ancient Egypt, despite the huge amount of textile preserved, it is possible that it was transported through another route than the more obvious contact through Egypt (Betts, 1994).
19
2.5 Wool
Wool is an animal fibre, which forms the protective fleece of several animals, especially sheep, goats and camels. It is composed of a protein, keratin and a few percentages of lipids, while vegetable fibres, such as cotton or flax, are mainly composed of cellulose. Protein substances are more exposed to chemical or environmental damages than vegetable fibres. Wool fibres are covered by small scales, which facilitate the adhesion of the fibres during the spinning and felting process, without needing the addition of other substances. They are also kinky, causing the formation of small air-pockets between kinks that give to wool good insulating properties (Barber 1991, 20). This structure makes wool flexible and elastic but has also high hygroscopy and a lower rate of flame spread. Wool fibres have a good compatibility with dyestuffs, which means that are definitely more easily to dye than vegetable fibres. Furthermore, wool occurs in nature in several hues, from black to white, giving the chance to create multicoloured textiles.
Quality of wool is generally stated by various factors, such as colour, crimps, yields and fibre diameter, which is the most important element. Generally, sheep’s fleece is made up of three elements: kemp, hair and underwool. Kemp is the outer and coarser part, with long and thick fibres but break easily. It is not possible, therefore, to spin them on their own. Hair and under wool are thinner, especially under wool, and can be spun either separately or together, depending on the type of yarn desired (Andersson Strand 2015, 42). Quality of fibre can depend from several factors, such as the origin of the fibres on the animal body (thighs, shoulder or side) and on the animal itself, ram, ewe or wether (wether being the best choice for wool).
2.5.1 First woollen textiles
Sheep original fleece was quite different from that of modern sheep. It was probably more similar to that of modern mouflon and, only with domestication and selection, it becomes woollier. Herding probably starts around 8500 BC in Taurus and spread to Levant and Mesopotamia gradually (Vila and Helmer 2014, 22). Neolithic communities exploited sheep breeding and select animals for meat, milk and fleece. Archaeozoological data, however, show that sheep were kept for their fleece starting only from 6500 BC (Vila and Helmer 2014, 23). Sheep and goats gradually became essentials in the economy of the Mesopotamian area and overwhelmed other breeding from the Uruk period onward. Wool was available to spinning only after the domestication of sheep, while plant fibres were exploited definitely earlier, as shown by flax findings. The selection of wool quality and the presence of sheep with uniform fleece, with few or total absence of kemp fibres, took place in different periods in the Mediterranean area. Samples from Tell el-Amarna (14th century BC) and Akrotiri (1645-1600
20 BC) show that it was already achieved in the Middle-Late Bronze Age in the Eastern Mediterranean (Gleba 2012, 3648).
In the Near East, first findings of woollen textiles are dated to the end of the fourth Millennium B.C. and come from the renowned site of Šahr-i-Soḫta in eastern Iran (period I) (Good, 1999, p. 110) and from Novosvobodnaya in the northern Caucasus. The Iranian textile is made from fibres of the Urial sheep (Ovis vignei) in two colours, yellow and brown, and probably preserve a starting border. The textile of the Caucasic Maikop culture is even more interesting. In fact, most of it is composed of wool, but also by a plant-fibre. The warp yarns occur in brown and beige colour while weft only in a beige colour and in both systems they are not plied. This forms a pattern of dark stripes on a light background (Shishlina, 2003, p. 333).
A fragment of a woollen textile comes from the Anatolian site of Arslantepe and was found in level VIB2 (Early Bronze Age I) in a royal tomb. It is of a fine quality and fibres are of goat hair and not sheep hair (Laurito, et al., 2014, p. 160).
First findings of wool for the Sothern Levant come from the tombs of Jericho (Shamir, 2015, p. 22; Crowfoot, 1960, p. 521) and do not predate Middle Bronze Age, definitely later than other regions. It might be possible that future excavations will predate the beginning of the wool exploiting at least to the Early Bronze Age, but certainly not for the Chalcolithic period, since the caves findings have shown that only vegetable fibres were in use.
Wool was certainly the predominant fibre in Mesopotamia, but our knowledge is based on textual sources rather than on actual archaeological finding. Archaeozoological findings in southern Mesopotamia testify a gradual decrease of usage of cattle and pigs and an increase of presence of sheep and goats for meat, dairy products and probably wool. Texts show a large-scale exploitation of wool from the half of the 3rd millennium onwards, but it starts to occur from the Late Uruk documentation. It was manufactured at all levels of society, from the domestic production to temple and palace workshops, from everyday usage to goods for gods and exports as a currency of exchange (Breniquet & Michel, 2014, p. 2).
There is little evidence of use of wool before the Roman Era in Egypt and they come scattered from all periods of Egyptian history. Unfortunately, no 14C test has been conducted on these specimens, so their chronology, deducted by archaeological contexts, can be challenged. The first findings come from the Neolithic site of el-Omari (DATA) and have been labelled as wool. Hairs of wool come from A 16 room C, but not directly studied by Greiss (1955, p. 228); n°15 from A 91 is a piece of sheep skin with wool of different colours (natural or not?) directly studied by Greiss (1955, p. 229)12.
12Unfortunately, Greiss’ report is extremely concise and does not offer any information about textiles except the definition
21 Another finding has a Predynastic date and comes from Naqada, Tomb 26 and it is described as a “brown and white woollen knitted stuff” (Petrie & Quibell, 1895, p. 24). In this case, the excavators exclude to be intrusive materials due to rats activities because of the difference with any other material found on the site and because of the depth of the finding.
Another finding dates to the First Dynasty and has been found at Helwan (Saad, 1951, p. 44). It is an actual cloth, which was wrapped around a body in tomb 36H5 but no other indications are given. Greiss (Greiss, 1955, p. 231) does not confirm the identification (if he even identified the correct textile) and writes “fibrous material associated with a coffin remains. Light and dark golden colour; human hair”. Another Old Kingdom specimen comes from the Pyramid of Menkaure at Giza, where a skeleton is recorded as being “enveloped in coarse woollen cloth of a yellow colour” but is generally considered as intrusive (Lucas & Harris, 1962, p. 147). Several examples of wool are known from Middle Kingdom Kahun and some from New Kingdom Amarna, as well as specimens of goat hair. One of the Kahun fragments, however, has been radiocarbon dated and was proved to be Roman (Vogelsang-Eastwood, 2000, p. 269).
48 fragments of sheep wool come from the workmen’s village of Amarna, 36 of which are definitely identified as wool, some of them from intact deposits. Other woollen objects come also from the Main City but have been excavated at the beginning of the 20th century (Kemp & Vogelsang-Eastwood, 2001, pp. 35-38).
Bones of ovine are regularly found in Egyptian excavations of all periods and they are frequently recorded in texts. Sheep and goat are also represented in several paintings, starting from rock art representation in Southern Egypt and Lower Nubia from at least Predynastic time and in some tombs of Pharaonic age. Two species of sheep are generally represented, one with long and spirally horns (Ovis longipes palaeoaegyptiacus) occurs in the oldest images, the other (Ovis platyura aegyptiaca) with curl horn, which probably provides a better quality of wool13.
Besides sheep and goat’s wool, Egypt has provided evidence also of camel wool from the Fayum Neolithic culture, where a rope was identified as made of this fibre (Caton-Thompson & Gardner, 1934, p. 88). Camel is not frequently used in Egypt, but there are a few hints, which prove their knowledge from the Egyptians. A figurine of a camel is known from Maadi, and camel bones were recognised in the site of Helwan (Saad, 1951, p. 38). Camel’s hair has not been recognized in the Near East, prior the 5th century BC, except for an uncertain identification from Šahr-i-Soḫta (Völling 2008, 73, Tab.1).
22 Even if certainly present in Egypt, so far very few specimens of wool have been found in Egypt. Tata (Tata, 1986, pp. 184-186) proposes that wool was probably sent abroad to countries which made a larger use of this fibre, but the most likely explanation continue to be that offered by Herodotus, who explains that wool was considered impure by ancient Egyptians. This element has led to an exclusion from the funerary contexts, which are our main source for the knowledge of Egyptian textiles, and have caused a substantial reduction of our chances to find woollen textiles. It is surprising, though, how strictly this prohibition was observed, since wool has not been found in any tomb of any age (Pharaonic of course) and of any social level. Moreno Garcia14 pointed out, however, that in many cases only tombs of a part of the society are excavated, like in the case of el-Bersheh, whereas the tombs of the poorer people have been compromised by water. In these tombs we could have had a better chance to find woollen textiles than in those of the Pharaonic court and of the élites, which were more aware of the religious prescriptions and closer to the priestly class. Furthermore, it is well known that a number of Asiatic people were present in Egypt, and at least for this part of the population we could expect to find some scraps of woollen textiles. Another possible explanation, which should be kept in mind, is that soil composition affects the preservation of materials and it is especially true for textiles. Basic environments destroy animal fibres, while vegetable fibres are not preserved in acid conditions (Gleba, 2011, p. 7). As suggested also by Kemp (Kemp & Vogelsang-Eastwood, 2001, p. 54), the dry climate of Egypt does not preserve every organic material in an equivalent manner. It is then likely that some soil with a basic PH might have helped to decrease the quantity of preserved woollen fabrics.
2.5.2 Wool processing
During the Bronze Age wool was probably obtained by plucking or cutting by knives. Only with the introduction of iron, it has been possible to produce good shears and to proceed to shear the animal. There is a remarkable difference between the two techniques, since cutting/shearing can be done twice a year, while plucking can be carried out just once a year, when the animal is moulting. Therefore, it is necessary to have the animals available when moulting, and this excludes that wool can be obtained without breeding.
After the wool is obtained, it can be sorted following several criteria, such as colour, fineness, crimp and length (Gleba 2008, 98). Wool can be washed before cutting or plucking, but in this case it is necessary to add a little fat to help the spinning process, which is not necessary if it is washed after
14Pers. comm. On wool consumption in Egypt see his speak during the workshop “Textiles in Ritual and Cultic Practices
23 spinning, because lanolin is still present. However, washing the sheep before the wool is extracted, makes the wool easier to card (Forbes 1956, 4:21). Washing the fibres is also necessary before the dying process, otherwise the dyestuffs will not be able to penetrate inside the fibres; generally, fibres are dyed before the spinning process.
Wool has to be prepared for spinning, teasing it by hand or combing in order to remove dirt from the fibres, untangle them and separate long fibres from the under wool. Combing or carding the fibres allow obtaining two completely different yarns, since the combed fibres produce a strong and hard yarn, while the carded ones a soft, elastic and fluffy yarn (Barber 1991, 20). However, it is not clear which tools were in use in the period under examination, since combs were generally linked to personal care (as seen further) and carding seem to appear in Roman times. The more carefully this operation is conducted, the easiest the spinning process will be and the yarn more uniform. As well as flax, wool discarded during these operations can be used for other purposes, such as padding or insulation (Andersson Strand 2015, 44).
24
3 From fibre to textile: spinning and weaving
Much has been written about spinning and weaving techniques15. This chapter aims to summarize the principal phases of these processes in order to better understand the archaeological materials and their function presented in the following chapters. For this reason, only a brief presentation of the general topic will be given.
3.1 Spinning
Spinning is the process which transforms fibres in a single, continuous yarn. The operation can be performed by using different methods and tools, and the result can differ considerably. The thread can be more or less even depending on the quality of fibres preparation, on the technique employed and on the ability of the spinner.
First threads and cordage were probably produced in the Upper Palaeolithic (Barber 1991, 39) and the first form of spinning was probably hand-spinning. In this method, fibres are simply drawn out from a mass and twist between hand-palms or palms and thigh; then more fibres are added and rolled again until the required thread is obtained (Carington Smith 1976, 69). However, this method is time-consuming and requires to wind the yarn around some objects in a separate step otherwise it will tend to untwist.
The following step probably consisted of the use of a spindle instead of the palms. A spindle could be a simple stick, which helps in the twisting of fibres and simultaneously allow winding up the yarn already produced. If a small weight – the spindle whorl - is added to the rod, the spindle will be able to rotate freely much longer. With this simple tool, spinning becomes much faster and the thread is much more even than that produced simply by hands. Furthermore, one hand is now free to control the provision of fibres and keep the spindle rotating. Spinning with a spindle has been performed until the spinning wheel was invented in the Middle Age (Carington Smith 1976, 70).
Several methods of spinning with a spindle have been used and Crowfoot classification (Crowfoot 1931, 9–20) distinguish between six methods16. For this work, it seems useful to explain the last three methods:
Grasped spindle spinning: a rove already prepared is passed through a ring or a fork and rotated with both hands. It is evaluated by Crowfoot as more suitable for doubling (Crowfoot 1931, 14)
15 For further and more specific readings see: (Barber 1991, 39–50; Crowfoot 1921, 1931; Kemp and
Vogelsang-Eastwood 2001, 74–81)
16 Hand spinning, spinning by twisting a hooked stick, rotation of spindle in hand, grasped spindle, supported hand spindle,
25 Supported hand spindle: drafting of fibres is a separate action. Spindle rests lengthwise on the
thigh. It rotates on the floor or in a cup. Very similar to suspended spindle
Suspended spindle spinning/drop spinning: the spindle rotates freely in the air and it can be either rotate by fingers or along the thigh to add momentum.
During the spinning process, the thread has to be wound around the shaft. With the drop spindle techniques, it happens when the yarn produced is so long the spindle reaches the ground. After the yarn has been wound up, the process can start again until the shaft is full.
Spindle whorl can be put equally on the top of the spindle or at the bottom (or even in the middle of the shaft). It is generally believed that the position of the spindle whorl on the shaft is culturally determined. However, some technical issues might influence this position. If the spindle is to be rolled along the thigh, a high-whorl spindle is preferred; if it has to be twirl between thumb and fingers, a low-whorl spindle might be preferred (Carington Smith 1976, 76).
It is generally believed, thanks to iconographic sources, that Ancient Egypt used a high-whorl spindle, while the Aegean-Anatolian area, as well as Europe, a low-whorl spindle. Mesopotamian and Levantine have fewer iconographic sources that can help in recognizing the spindle whorl position and it will be discussed in the following chapters.
As seen in the previous chapter, wet fibres have a natural twisting orientation, which is s for flax and nettle, z for hemp, while cotton can rotate in both directions but give best results when is z-spun (Forbes 1956, 4:151). Animal fibres, like wool, do not show an orientation when moistened and can be spun in both directions. The natural orientation, however, might help in the process of spinning but it is not binding; all fibres can be spun in both directions. To prevent the yarn from untwisting and to add strength, the thread is generally doubled or plied, which means that an extra yarn with the same twist of the first is added, and the two are then spun together, generally in a contrary direction to that of the single yarns. The directions are called s (counterclockwise) and z (clockwise), from the central part of the letter, which resembles the twist lines
visible on a yarn. It means that if two threads are s-spun, they will be plied in a Z direction (conventionally Z, 2s). When plied threads are twisted together, it is called cabling. Analysing of ancient textiles, however, show very different adaptation of thread directions. Thread quality is determined by its thickness and by being tightly or loosely
spun. The angle of the twist determines the tightness and it can be defined as loose if it is less than 25°, medium if it is between 25 and 45°, or tight if it is more than 45° (Jones 2002, 324).
26 First flax threads show, as seen in the previous chapter, that the first direction of spinning used was z, and the plying direction was its contrary, meaning S. It worth noting that this is true both for textiles from the Near East as well as Egypt, as shown in tab. 1. During the Chalcolithic period, it is possible to assist to a change in the spinning direction, with a prevailing of s-spun and S-plied threads at least in the Levant (Schick 2002, 223–24).
Table 3 Spinning directions in first textiles
Slowly, also in Egypt flax textiles characterized by the combination of s-spun and S-plied threads, seem to prevail. Thanks to the major quantity of preserved textiles it is clearly visible a transitional phase in the Egyptian Predynastic period. It possibly begins at the end of the Naqada I phase but it certainly continues during the Naqada IIA phase, while with the beginning of the IIB phase textiles show the complete adoption of s-spun, S-plied solution (Jones 2008, 109, 111). This structure will be prevalent for all the subsequent Pharaonic phases, and a Z-twisting will be found only to a lesser extent. Textiles generally show an S, 2s structure of threads, both for warp and weft, but occasionally z-twist can be observed on nets and sewing threads (Kemp and Vogelsang-Eastwood 2001, 60). In Amarna textiles, for example, the s direction for spinning and plying covers the 99% of the entire set of samples (Kemp and Vogelsang-Eastwood 2001, 59). The s-twist seems to be particularly suitable for spinning flax, because it follows the natural orientation of the fibres. It is curious to observe that also Amarna wool samples show the same structure, S, 2s (Kemp and Vogelsang-Eastwood 2001, 60). It might be explained with the spinning technique prevailing in Egypt, meaning the drop spindle. A right-handed person will tend to roll the spindle along the thigh from the hip toward the knee, producing an s-spun thread, made of flax or wool indifferently. However, if the momentum is not added by rolling the spindle on the thigh of the spinner, but simply by the use of thumbs and fingers, like a top, the spinning direction will be this time the exact contrary. This type of movement, however,
Wadi Murabbaʽat z-spun, S-plied (Schick 1995, 201)
Na al emar cordage Z-plied
nets z-spun, S-plied; fewer s-spun, Z-plied examples of cabled threads (Shamir 2014, 143)
Çatal Höyük z-spun, S-plied (Burnham 1965 p. 171)
Fayum fragment z-spun, S-plied (Jones 2008, 107)
Badarian textiles z-spun, S-plied (Brunton and Caton-Thompson
1928, fig. lix)
Nahal Mishmar and other Chalcolithic caves s-spun, S-plied (Schick 2002, 231) Predynastic textiles Transition from z to s (Jones 2008, 111)
27 is generally used with spindles with low-whorls, while the use of the thigh is preferred with high-whorl spindles (Carington Smith 1976, 78–79).
Ancient Egyptian and probably Levantine people seem to not have used a distaff to keep fibres in order before they were added to the spindle. For keeping parallel long fibres such as those of flax they develop another method which is called splicing. It consists by joining the single fibres together to form a long roving, which can be coiled to form a ball of yarn and only subsequently it is actually spun. More specifically, the bundles of the fibres are overlapped for a dozen of centimetres and this part twisted by rolling it by hand, on the thigh or on another surface. Some glue substances may be added to reinforce the joining point, such as saliva, but also water might be sufficient to free the natural glue present in the flax.
Splicing has been recognized both on Egyptian and Levantine textiles (Gleba and Griffiths 2011, 287) from the very beginning of their history, and continue to be performed in Egypt until probably the Roman conquest. In the Levant, it seems to disappear during the Bronze Age (Shamir 2014, 147). Other cultures have developed the same technique outside the area under consideration, such as Central Europe (Leuzinger and Rast-Eicher 2011, fig. 3 d; Gleba and Harris 2018).
Table 4 Occurrences of spinning and weaving scenes in the Egyptian iconography
REPRESENTATIONS OF SPINNING AND WEAVING SCENES
Bowl from el-Badari Bowl with representation of a horizontal loom. On the side, two figures beside vertical poles might have been engaged in warping (Cortes 2011, 94)
Saqqara, VI dynasty Relief with a man with a spindle on one hand (note the conical spindle whorl!) and a ball of yarn under left leg (Firth and Gunn 1926, 36; Barber 1991, 76).
Tomb of Baqt, Beni Hasan (MK) Upper register: 3 men spinning, one with grasped spindle, two drop spindle. One man preparing a net, two men preparing balls of yarn.
Lower register: 4 women couching on the floor preparing roves; 3 spinners, each with two spindles (drop spindle) and two spinning bowls; 7 women attending two looms (Percy Edward Newberry and Griffith 1893, fig. IV)
Tomb of Khety, Beni Hasan (MK) Upper register: 3 men spinning, one with grasped spindle, two drop spindle. Two men preparing a net, one man matting. Lower register: 4 women couching on the floor preparing roves; 3 spinners, each with two spindles (drop spindle) and two spinning bowls, 5 in total because they share the same bowls; 7 women attending two looms.
(Percy Edward Newberry and Griffith 1893, fig. XIII) Tomb of Khnumhotep, Beni Hasan
(MK) A spinner with one drop spindle and two yarns, two bowls and a crouching woman preparing fibres. A horizontal ground loom with two female weavers and a male supervisor.
28 (Percy Edward Newberry and Griffith 1893, fig. XIX)
Tomb of Daga, Thebes (MK) Upper register: a woman warping, two women crouching and attending to a horizontal loom.
Lower register: 3 women preparing fibres, one attending to a spinning bowl and another spinning with two drop spindles (Davies 1913, fig. XXXVII).
Tomb of Djehutyhotep, el Bersheh
(MK) Upper register: 5 women couching on the floor preparing fibres/roves. 3 spinners with one spinning bowl each. One woman attending warping on three pegs.
Lower register: 2 women pulling threads from boxes of balls of yarn (?). 3 women warping a loom.
(Percy E. Newberry and Griffith 1894, fig. XXIV) Tomb of Sarenput I, Elephantine,
(MK) Upper register: one person couching on the floor preparing fibres/roves. Two spinners with two spinning bowls each. (Müller 1940, 41)
Tomb of Neferrenpet, Sheikh Abd el-Qurna.
Chief of the weavers in the Ramesseum (Ramesses II)
Two people couching on the floor preparing roves. Two warping frames with an attending person each.4 vertical two-beam looms with a total of 5 weavers (men and a woman). Upper register in a very bad state, a man, 4 women and ½ children are employed. Scale with coils of thread and raw material (white and yellow).
(Davies 1948, fig. XXXV)
Neferhotep, Thebes, 18th dynasty Representation of temple estate: A vertical two-beam loom with one operator and probably a second one on the right, with a second weaver.
(Davies and Davies 1933, 38, pl. XLIX) Tomb of Djehutynefer, Thebes,
18th dynasty Two people preparing fibres, two preparing coils of yarn. Grasped spinning scene, one spinner, two spindles fixed two a ring each and both rotating in a bowl. Two vertical two-beam looms with a total of 3 weavers
(N. de Garis Davies 1929, fig. 1a) Ostracon from Deir el-Medina
(NK) Grasped spinning scene: two spindles hanging from a ring. (Vandier d’Abbadie 1937, fig. LXIV) Funerary papyrus of Dirpu, 21th
dynasty, Deir el-Bahari The deceased is taken by hand by a cat-headed goddess and in the other hand carries a high-whorl spindle with dome shaped spindle whorl and thread wound around the shaft. (Piankoff 1957, fol. 7)
FUNERARY MODELS
Weaving model, tomb of Meketra,
Thebes, 11th dynasty Three couched women preparing fibres; three spinners with two spindles and a spinning bowl each; two women warping on the wall with a spindle each; two horizontal ground looms, with three weavers in total.
(Winlock 1955, 29–33, pls. 26–27, 66–67) Weaving model tomb of
Anpuemhat and Usernekhbet, Saqqara (MK)
Two people couching on the floor preparing fibres/roves. A spinner with two spindles and a spinning bowl. A horizontal ground loom with two weavers.
(Tata 1995, 133–36, fig. 16; Kemp and Vogelsang-Eastwood 2001, 325)
29 Weaving model of Gemniemhet,
Saqqara (MK) Two people couching on the floor preparing fibres/roves. A spinner with two spindles and a spinning bowl. A horizontal ground loom with two weavers on their left. On the other side less clear part with a standing spinner, two people preparing fibres and three people working on a loom. (Firth and Gunn 1926, 52–54, pls. 22b, 31c)
Model from Girgeh (MK) A woman preparing fibres, one spinner, one woman warping on a warper fixed to the ground, two weavers.
(Tata 1995, 129–30; Hayes 1953, fig. 172) Model from tomb of Khetya, Beni
Hasan, Tomb 575 (now Liverpool) (MK)
No walls represented. A spinner with two spindles, one with high whorl and the other with low whorl, but originally both were high-whorl. Spinning bowl is now lost. Two weavers work on a horizontal ground loom (only poorly preserved) (Garstang 1907, 133; Tata 1995, 136–38)
Sedment tomb 1525 (MK) No walls represented. Standing spinner (yellow skin) and crouching figure (red skin)
(Petrie and Brunton 1924, fig. 17; Allen 1997, 31)
Deir el-Bahari, Montuhotep (MK) Possibly model of textile workers: two men with raised leg (spinners?), rods which might have belonged to the model. (Arnold 1981, 33, pl. 50a, 58d; Allen 1997, 31)
Model from tomb of
Djehutynakht, el-Bersheh 10A (MK)
No walls represented. A spinner with one spindle, a crouching person attending the spinner. Two weavers work on a horizontal ground loom (only poorly preserved).
(Allen 1997, 31) Boston MFA 21.891.
Unknown, Munich Model of a man with raised leg, possibly a spinner (Breasted 1976, 53, pl. 47c; Allen 1997, 31)
Unknown, Berlin Model of a weaving room, with remains of a standing figure and four squatting figures. (Johl 1924, 9–10, Abb. 41-43; Allen 1997, 31)
Karenen and Nefersemdet,
Saqqara (MK) Model of a room with arched beams which possibly represented a weaving workshop. No figures or implements preserved (Breasted 1976, 53; Allen 1997, 32)
Karenen and Nefersemdet (MK) Model of a room with arched beams which possibly represented a weaving workshop. Five figures preserved, one with right leg raised (Breasted 1976, 54; Allen 1997, 32). Djehuty, el-Lisht Workshop with granary, bakery and weaver’s shop. One
woman preparing fibres, one probable spinner with spinning bowl and two weavers.
(Allen 1997, 31) Indiana Univ. Art Museum, 58.34.
Iti, Lahun, tomb N17 Small model of a spinning bowl with a hole in the bottom, where a peg to attached to the model should be inserted. Petrie museum UC 6665i (Petrie, Brunton, and Murray 1923, 33– 34)
Unknown Lisht, tomb 428 Model of limestone spinning bowl (Allen 1997, 31). MMA 15.3.99
30
3.2 Spinning tools
As seen previously, a spindle is a simple rod to which a weight might be attached to extend the rotatory movement. However, recognising what actually is a spindle or a spindle whorl in an archaeological context might be difficult. Spindles can be confused with pins, for hair and dress, or with kohl-sticks. Spindle whorls can be confused with beads, and in many cases it is impossible to define what is a spindle whorl and what is a bead. For this study, objects are generally classified following the indications of the excavators, because they had the opportunity to handle them and to analyse them in context with the other findings of the site. However, in some cases, the original indication seemed not correct and objects have been reconsidered.
3.2.1 Spindles
Spindles can be made of several materials, primarily wood, but also luxurious examples were produced, made of metal, ivory and bone. Given the organic nature of the everyday use tools, almost all the ancient spindles have disappeared, except in some specific contexts. Most of the spindles known from all over the world are then the precious specimens and are quite rare to find.
Even if they could be of different shapes and different length (and a look to modern spindles could be a good exercise in that sense), it seems useful to trace some lines which could help in recognizing an item as a spindle.
Length is certainly an important parameter in a spindle since it determines the ease of its use. A spindle should be long enough to be easy to manipulate, but also to accommodate the spun yarn in order to avoid that it unspins. However, it should not be too long or will be complicated to twist and will reach earlier the ground (for drop-spinning) producing only small quantity of yarn before needing to be wound up again. In her work on pre-roman Italian spindles, M. Gleba (2008, 103) reports that local spindles are generally 30 cm long, but this length is reduced during the Roman period, when spindle’s length ranges between 12 to 27 cm. Some exceptions are also present, with objects reaching 40 cm. Length is probably related to the weight of the spindle whorl and thus to the fibre in use and the final product desired.
The shape of the spindle can vary greatly as well. A spindle can be a simple cylindrical rod, or could have a thick central part and thin ends. Generally, it tends to tapers toward one or both the extremities, and has the thicker part at one-third of the length, while for the other two thirds the shaft gently becomes thinner toward the tip, which could be either rounded or pointed.
Other features might help in recognizing spindles. The first and more easy to identify is a notch or a groove in the upper part of the shaft, where the yarn can be fastened to make easier the process of
